LANDSLIDE RESEARCH:
How much is enough?
By David Hockman-Wert
After more than 30 years, a hundred research
studies, and who knows how many thousands of dollars, scientists are still
trying to answer a fairly straightforward question: does clear-cut timber
harvesting cause landslides? While it may seem odd that a conclusive
answer has not been reached, the issue of causation is not as clear-cut as
some of the ground being studied.
In the past few months, Oregonians have seen a new
twist in the battles
over forest practices in the state. Previously, environmental groups
argued that clear-cutting damages sensitive and valuable forest
ecosystems. Timber companies counter that
society needs wood, and as long as harvesting is done carefully, they will
continue to fulfill these needs. Since the landslides of November 1996,
the pitched battle has moved to a new front: the threat clear-cutting
poses to public safety.
Landslides originating in recent clear-cuts have
demolished homes, covered roads, and killed people. A ballot initiative to
ban clear-cutting finds this to be sufficient evidence to convict
clear-cutting of "causing landslides." Yet landslides have been occurring
for hundreds and probably thousands of years in Oregon, before
clear-cutting -- or any sort of timber harvest -- was widespread.
In the midst of the public argument, research
studies are often cited or refuted. Tim Hermach, director of Native Forest
Council, cites the results of three studies and declares with no ambiguity
that "the problem is not 'landslide-producing storms,' but
landslide-producing logging." On the other hand, State Forester James
Brown claims that "earlier studies ... don't provide us with complete
answers."
What is the public, generally not well-versed in
the geological intricacies of landslide science, supposed to think? For
Hermach, it is clear: "The public should understand that the facts are not
in dispute here. ODF [Oregon Department of Forestry] and the governor
simply don't like the facts ..."
Are the facts not in dispute? Is the state
forestry
agency burying the "facts"? What do the studies say?
Scientists have studied the causes of landslides
since the mid-1960s. Nearly every research study indicates that clear-cut
timber harvesting increases the likelihood of landsliding on steep and
unstable hillsides (from two to 31 times) and the erosion caused by
landsliding (from two to 41 times), compared with uncut forests. These
results appear to show general agreement that clear-cutting increases the
incidence of landslides. However, some forest scientists still aren't
convinced of this. Considering the methodology of some of the studies and
the harvest practices of an earlier era, these scientists may have good
reason to be skeptical.
A Brief History of Landslide Research
The first study to identify timber harvesting's
effects on landslide rates was done in 1949 in Utah. Two U.S. Forest
Service research scientists, Croft and Adams, found that heavy rains and
snowmelt initiated landslides, but they also thought that timber cutting
and slash burning contributed to reducing the stability of the hillsides.
In the 1960s, after easy-to-log areas had already
been harvested, and steeper and more unstable sites began to be logged,
concern arose about logging's impact on soil erosion. In 1972 the Forest
Service initiated the Interstation Soil Mass Movement Research Program, to
focus on the "causes, effects, prevention, and impacts of soil movement."
Corvallis, Oregon, is home to one of the three stations in the program.
C.T. Dyrness, a Forest Service soil
scientist
working at the Corvallis research station, and Fred Swanson, a Forest
Service geologist then at the University of Oregon, examined air photos
and on-the-ground studies of the western Cascades in 1975. In their
oft-cited paper, they noted that landslides occurred three times more
frequently in clear-cuts, and carried roughly three times as much debris.
Notable, however, is that all of this soil
movement happened in an "unstable zone," a lower elevation area with older
unstable soils. In a geologically "stable zone" of roughly equal area, no
slides occurred from either forested stands or clear-cuts.
These findings were supported by a 1979 study of
the effects of a major storm in 1975 on the Mapleton Ranger District of
the Siuslaw National Forest. The Mapleton study found that 95 percent of
all slides occurred on slopes steeper than 70 percent (35 degrees) and on
three of the most unstable soil types. These studies demonstrate the
scientific consensus that the steepness and inherent instability of a site
play key roles in the initiation of landslides.
Mapleton researchers also found 8.5 times more
slides in clear-cuts than in forested areas, and 21 times the volume of
debris. They caution, however, that these data do not include all of the
slides in forested areas: "an unknown number, occurring at inaccessible
sites or too small to see under heavy forest cover, were missed."
Two well-respected ground-based studies were
performed in 1977, also on the slide-prone Mapleton Ranger District. In
the first, Gary Ketcheson, a graduate student in the School of Forestry at
Oregon State University (OSU), found that the frequency of medium and
large slides in clear-cuts and forests were roughly the same: one slide
per 21 acres in clear-cuts compared to one slide per 28 acres in uncut
forests. The erosion rate from clear-cuts, however, was significantly
higher (3.7 times) than in uncut forests.
In the second study, a team led by Fred Swanson
found that the frequency of slides in clearcuts was actually somewhat
lower than in the forest: one slide per 44 acres in clearcuts and one per
31 acres in uncut forests. Yet the erosion rate from clear-cuts was twice
as high as from uncut forests.
These studies illustrate the dilemma of drawing
conclusions from study
results. If only one of the steepest and most unstable soil types is
considered, the frequency of slides in clearcuts becomes 3.5 times higher
than in forests for Ketcheson's study: one slide per 11 acres in
clear-cuts compared with one slide per 38 acres in uncut forests.
Swanson's results also show an increase in slide frequency from
clear-cuts: one slide per 24 acres in clear-cuts and one slide per 31
acres in uncut forests. Erosion rates likewise increase, to 12 times
higher in clear-cuts for Ketcheson and to four times higher in clear-cuts
for Swanson.
Perhaps the most compelling paper of this era was
a catalog of 43 landslide inventories prepared by George Ice for the
National Council of the Paper Industry for Air and Stream Improvement. The
paper concluded that "the increase in mass soil movement due to
clear-cutting varies widely, ranging from 2-4 times in Oregon and
Washington."
A tentative consensus
Volumes were written explaining how timber
management activities, especially roadbuilding, exacerbated the tendency
of steep and unstable
hillsides to fail and slide down the slope. Douglas Swanston and Fred
Swanson, geologists working in the Forest Services' Pacific Northwest
Experiment Station, wrote, "Timber harvesting operations, particularly
clear-cutting and road
construction, accelerate these [landsliding] processes, the former by
destroying the stabilizing influence of vegetative cover and altering the
hydrologic regime of the site ... "
In other words, when the roots of dead trees lose
their
strength, and water saturates the ground and runs off in higher
quantities, the risk of
landsliding increases.
At a 1984 workshop on slope stability and forest
management, F. Dale
Robertson, then associate chief of the U.S. Forest Service, declared what
was becoming a common understanding among forest land managers: "We should
understand that now and then our activities will accelerate or trigger
soil movement."
Recent Landslide Studies
The landslide issue quieted for more than a decade as
the weather in Oregon hit a lull. From 1981 to 1996, storms were mild and
did not cause significant landsliding.
In February 1996, the most intense storm since
1964 hit the Oregon Coast Range. In the wake of the February floods,
interested groups commenced a flurry of studies. Fifteen years of forest
management was put to the test.
First to be reported were the aerial surveys. The
Association of Forest Service Employees for Environmental Ethics (AFSEEE),
an environmental group based
in Eugene, flew over the Mapleton Ranger District. They found that 62
percent of the slides they saw originated in clear-cuts compared with only
two percent of slides occurring in the forest. The remainder were
road-related.
Pacific Watershed Associates, a geological
consulting firm from Arcata,
California, performed their own aerial survey over large areas of the
Coast Range and the Cascades. Their findings corroborate AFSEEE's Study:
71 percent
of located slides were in recent clearcuts, compared with six percent in
uncut forest.
Siuslaw National Forest staff conducted their own
fly-over survey and
found similar patterns of landslide occurrence.
"Ground-truthing" aerial study results
Meanwhile, in early February, the Oregon Department
of Forestry (ODF) began its own massive study in conjunction with OSU,
Oregon Department of Fish and Wildlife, and the Oregon Department of
Transportation. Like the other studies, research began with aerial
observation. Unlike the other aerial surveys, ODF researchers followed up
on the photos by walking all the creeks in the study areas.
"Ground-truthing," as it's called, allows researchers to locate smaller
slides which go undetected under the forest canopy.
It would be hard to find another scientific study
which has been so anticipated and lauded before it's even completed. "I'm
very excited to apply their findings," says Jim Furnish, Supervisor of
Siuslaw National Forest. "[The study is] the most detailed assessment to
date, which will prove very useful."
This is the "only research of its kind in terms of
the amount of ground-truthing," notes Jim McCauley, director of the State
Timber Purchaser's Division of the Oregon Forest Industries Council.
Leslie Lehmann, executive director of the Oregon
Forest Research Institute, says the study is "really important,"
particularly because it is the first major study to analyze the
effectiveness of recent amendments to the Forest Practices Act. In other
words, the ODF study will test whether the present rules are sufficient to
protect forest health and public safety from landslides.
The ODF study is in the analysis phase; final
conclusions will not be drawn until late 1997. Yet the preliminary
results, according to Keith Mills, the ODF geotechnical engineer in charge
of the forest practices section of the study, indicate that "there can be,
on the steepest slopes, an increase in landslide risk in the first ten
years" after clear-cutting.
The studies performed in the wake of the 1996
flooding have revived the idea that logging causes landslides. Three of
the four studies show a much greater incidence of landslides in clear-cut
areas compared with forested areas, while one indicates a meager increase.
The consensus appears to be complete; even members of the timber industry
agree that forest practices can increase the risk of landslides.
"Nobody's going to deny that during the first ten
years after harvesting, you do have a greater risk [of landsliding]," says
Rex Storm, forest policy analyst with Associated Oregon Loggers.
Points of dissension
But not everyone is convinced.
"There's no real hard data that says we know that
there's an increase," says Arne Skaugset, a forest engineering professor
at OSU. "It's really hard to make the case that you're not increasing the
risk somewhat," but researchers just don't have the hard-core scientific
data to be sure about this.
Skaugset agrees that the studies show a marked
increase in the frequency of landslides from clear-cuts. "It is possible
to go into the literature and see that timber
harvesting is horrible," he says. But he claims that the data-gathering
techniques
can have as much effect on the results of a study as the harvest
techniques being tested. And he doesn't necessarily believe that results
showing a small increase in slide frequency conclusively prove that an
increase exists. Too many variables exist in landslide studies which could
cause the results to be slightly inaccurate. "I give the system a lot of
fuzz, since I think it's a fuzzy system," says Skaugset.
Swanson's and Ketcheson's studies, for example,
showed relatively small differences in landslide frequency between
clear-cuts and uncut forests. But the studies covered fairly small areas.
Keith Mills notes that differences in storm
intensity across a landscape might skew the results of such a study. If a
small study area receives considerably more rainfall than the surrounding
area, or vice-versa, the study results could be quite different from the
bigger picture. Larger landscapes need to be studied in order to guard
against misrepresenting the whole.
Two main issues cause scientists to disagree about
the applicability of research study results. One is the methodology used
by some studies, particularly aerial photos and surveys. The other issue
is that the type of harvest and road-building techniques being studied are
no longer used in the forest.
Limitations of aerial studies
Aerial surveys show the greatest difference between
the number of landslides in clear-cuts compared with uncut forests. This
methodology is quite common in the field of landslide research.
"Probably 90 percent of the studies out there are
based on aerial studies or windshield surveys [driving forest roads and
counting landslides]," notes Marvin Pyles, Professor of Forest Engineering
at OSU.
Such surveys can lead to "lower detection success"
of landslides in forested areas, says Swanson, meaning that aerial surveys
miss slides under the cover of trees. Some people hoped that was a "fatal
flaw" in the theory that clear-cutting increases landsliding. However,
Swanson warns against distrusting all landslide studies, especially those
which are ground-based. Most of those studies show a slight increase in
the likelihood of landsliding after clear-cutting, especially on the
steepest and most unstable slopes.
Andy Stahl, director of AFSEEE, claims that the
aerial studies his group
performed are "legitimate with the proper caveats which I included in my
study." In the study, he acknowledges that the number of natural slides
is "somewhat greater than recorded," but even so, he asserts that "the
vast bulk of sliding is logging-related." Besides, he says, the small
slides in the forest hardly ever lead to debris torrents.
The preliminary reports from ODF's study indicate,
however, that aerial studies are deficient in locating many landslides.
More than 50 percent of the ground-truthed landslides could not be seen on
the aerial photos.
"It doesn't take a brain surgeon to realize that
if you're looking out of an airplane, those slides hidden under the canopy
won't be visible," says Rex Storm of Associated Oregon Loggers.
In a three-square-mile stand of 200-year-old trees
where ODF's and the Forest Service's study areas overlapped, ODF's
ground-based methods located 40 landslides. The Forest Service's air
photos had not located any of these slides. It is results such as these
which cause researchers to be skeptical about the conclusions drawn from
aerial studies.
Older studies and outdated techniques
Besides the limitations of aerial surveys, the
results of older studies have been called into question because they are
based on logging and road-building techniques that haven't been used for
many years. The studies completed in the 1970s considered the effects of
harvest practices from the 1950s.
"The activity that took place then was pretty
rudimentary," says the OFIC's Jim McCauley. Logging with tractor skidders
compacted and disturbed the soil. Large quantities of post-logging slash
was burned at very high temperatures, removing shrubs and the organic
layer of soil.
In recent years, though, soil disturbance has been
minimized on steep slopes by using cable logging -- which collects logs
without a tractor skidder -- and by reducing the amount of slash burning
to retain shrub root systems. Less ground disturbance leads to some
decline in erosion rates, but it is unclear whether it significantly
reduces the frequency of landslides.
The issue of timing
Another issue of contention in the research community
is the timing of
landslides. Even if clearcutting increases the risk of landslides
in the first ten years after logging, it may not increase the number of
slides over the long term. If a site is prone to sliding anyway,
clear-cutting hastens the inevitable but doesn't cause more slides.
ODF's study supports this theory. Mills says, "there is an increase in
frequency [of landslides] sometime in the first ten years [after harvest],
and then a decrease."
Is root strength the primary component of soil stability?
Disagreement continues over one of the potential
causes of landslides: root decay. Some question the assumption that root
strength is the main factor in maintaining soil stability -- after all, if
roots are the primary mediator of soil stability, forested sites should
never slide. Likewise, harvested sites with decaying roots should never
hold. But as Skaugset points out, a site "can fail with the roots in it."
Most landslide specialists still consider root
strength to play a role in soil stability, but the precise role is
unclear. Skaugset explains that while it may be relatively easy to
identify landslide-prone sites, it is not easy to identify where a
specific slide will occur. Scientists don't know enough about the process
by which slides occur.
"When you go from landscape-level studies to
process-level studies, it gets truly gnarly," says Skaugset.
Headwall leave areas: do they work?
In the race to solve the landsliding problem, the
Siuslaw National Forest developed the headwall containment area program in
1975. Harvest plans under this program require leaving an area of uncut
forest around the highest-risk section of a
logged area. While headwall leave areas were intended to keep the
high-risk sites from sliding, Swanson says it was "never demonstrated that
they were effective." Recent inventories of headwall leave areas in the
Mapleton Ranger District have shown that landslides are as frequent in
these areas as in clear-cuts or uncut forest.
Nevertheless, Supervisor Jim Furnish maintains
that headwall protection is still prudent. "There's every advantage to
have a forested headwall as opposed to a bare headwall," he says. If a
forested headwall fails and a slide goes, the debris will include large
logs and branches which may improve in-stream fish habitat. A bare
headwall, in contrast, would dump soil and rocks into the stream, silting
up the gravel beds vital for fish spawning.
AFTER 30 YEARS OF LANDSLIDE RESEARCH, scientists
still don't know the precise causes of landslides -- or what role, if any,
timber harvesting plays. Why don't researchers know more after all the
studies that have been done?
As Marvin Pyles puts it, "The earth doesn't
cooperate in our statistical studies." The complex interactions of water,
rock, soil, and vegetation elude science's full understanding. This
complexity is enough to lead Keith Mills, the head researcher in ODF's
landslide study, to claim, "We don't have a lot of information. What we do
have, we still don't know what it all means."
Keith Mills says the two key questions are:
- "What role does vegetation have on landslide incidence?"
- "What role does vegetation have on the substances that debris flows
move?"
The problem with science's slow slogging through
complex and ambiguous matters is easily seen in the remnants of a
mud-ravaged home.
"People don't have much patience for complexity,
especially when houses are sliding down the hill," says Leslie Layman of
the Oregon Forest Research Institute.
Policy can't wait, though, for 100 percent
certainty; it must act within the ambiguity.
Commenting on the highly-charged political
dialogue, Swanson says scientists and politicians can continue to bicker
about whether the cutting should stop, but the real issue is to pay more
attention to what is happening and what has happened on the land.
"Rather than trying to stop landslides," he says,
"we should figure out how to put up with them."
SOURCES:
- Association of Forest Service Employees for Environmental Ethics.
2/14/96. Aerial Landslide Survey of Mapleton Ranger District Following
Rainstorm of February 1996.
- Brown, J. 2/6/97. The Problem of Landslides. Eugene Weekly.
16(5):8.
- Croft, A.R. and J.A. Adams. 1950. Landslides and Sedimentation on the
North Fork of Ogden River. May 1949. USDA Forest Service Intermountain
Forest & Range Experiment Station Research Paper INT-21.
- Dent, L., G. Robison, and K. Mills. 1996. Oregon Department of Forestry
1996 Storm Impacts Monitoring Project. Unpubl. Rep. Oregon Dept. of
Forestry.
- Furnish, J. 2/26/97. Supervisor, Siuslaw National Forest. Personal
communication.
- Gresswell, S., D. Heller, and D.N. Swanston. 1979. Mass Movement
Response to Forest Management in the Central Oregon Coast Ranges.
USDA Forest Service Resource Bulletin PNW-84.
- Hermach, T. 2/6/97. Look at the Evidence. Eugene Weekly. 16(5):
9.
-
Ice, G.W. 1985. Catalog of Landslide Inventories for the Northwest.
New York: National Council of the Paper Industry for Air and Stream
Improvement Technical Bulletin 456.
-
Ketcheson, G.L. 1978. Hydrologic Factors and Environmental Impacts of
Mass Soil Movements in the Oregon Coast Range. M.S. Thesis, Oregon
State University, Corvallis, OR.
-
Lehmann, L. 2/17/97. Executive Director, Oregon Forest Research Institute.
Personal communication.
-
McCauley, J. 2/17/97. Director of State Timber Purchaser's Division,
Oregon Forest Industries Council. Personal communication.
-
Mills, K. 2/25/97. Geotechnical engineer, Oregon Department of Forestry.
Personal communication.
-
Mills, K. 1996. Landslide Occurrence and Forest Practices (Exclusive of
Road Management) Associated with the Storm of 1996. Unpubl. Rep.
Oregon Department of Forestry.
-
Pyles, M. 3/5/97. Professor of Forest Engineering, Oregon State
University. Oregon Board of Forestry meeting presentation.
-
Robertson, F.D. 1985. Landslides on Steep Forested Terrain: The Problem
and Resource Mangement Implications. Keynote address in Swanston, D.N. ed.
Proceedings of a Workshop on Slope Stability: Problems and Solutions in
Forest Management. USDA Forest Service General Technical Report
PNW-180.
-
Skaugset, A. 2/28/97. Assistant Professor of Forest Engineering, Oregon
State University. Personal communication.
-
Stahl, A. 2/24/97. Executive director, Association of Forest Service
Employees for Environmental Ethics (AFSEEE). Personal communication.
-
Storm, R. 2/27/97. Forest Policy Analyst, Associated Oregon Loggers, Inc.
Personal communication.
-
Swanson, F. 2/25/97. Geologist, Siuslaw National Forest. Personal
communication.
-
Swanson, F.J. and C.T. Dyrness. 1975. Impact of Clear-cutting and Road
Construction on Soil Erosion by Landslides in the Western Cascade Range,
Oregon. Geology 3(7): 393-396.
-
Swanson, F.J., M.M. Swanson, and C. Woods. 1977. Inventory of Mass
Erosion in the Mapleton Ranger District, Siuslaw National Forest.
Unpubl. Rep. on file, Forestry Sciences Lab. Corvallis, OR.
-
Swanson, F.J., M.M. Swanson, and C. Woods. 1981. Analysis of
Debris-avalanche Erosion in Steep Forest Lands: An Example from Mapleton,
Oregon, USA. Erosion and Sediment Transport in Pacific Rim
Steeplands. Proc. Christchurch symposium.
-
Swanston, D.N. and F.J. Swanson. 1976. Timber Harvesting, Mass Erosion,
and Steepland Forest Geomorphology in the Pacific Northwest. pp. 199-221
in Coates, D.R. ed. Geomorphology and Engineering. Stroudsburg, PA:
Dowden, Hutchinson & Ross, Inc.
-
Weaver, W. and D.K. Hagans. 1996. Aerial Reconnaissance Evaluation of
1996 Storm Effects on Upland Mountainous Watersheds of Oregon and Southern
Washington. Arcata, CA: Pacific Watershed Associates.
|
